can you help us for area segmentation method on temporal bone object to segment the region marked with arrows

can you help us for area segmentation method on temporal bone object to segment the region marked with arrows

Thank you

Answer 1

[Image Analyst]

Did you try thresholding? See attached demo. Adapt as needed.

Answer 2

[yanqi liu]

sir，please check the follow code to get some information

for more information,you can read the book 《计算机视觉与深度学习实战》，Thank you！

clc; clear all; close all;
im = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/751599/image.jpeg');
im = rgb2gray(im);
% make bw
bw = imbinarize(im,'adaptive','ForegroundPolarity','dark','Sensitivity',0.85);
% choose target
bw2 = imfill(bw, 'holes'); bt = imclose(bw2, strel('disk', 2));
bw2 = imfill(bt, 'holes');
bw2 = imopen(bw2, strel('disk', 10));
bw2(:,round(size(bw2,2)/2):end)=0;
bw2 = bwareafilt(logical(bw2),1);
bt = logical(bw.*bw2);
bt2 = ~bt; bt2 = imclearborder(bt2);
bt(bt2) = 0;
bt = logical(bt); 
bt = bwareaopen(bt, 80);
bt = imclose(bt, strel('disk', 20));
bt(bt2) = 0;
bt = logical(bt); 
% make result
im1 = im; im2 = im; im3 = im;
im1(bt) = 255; im2(bt) = 0; im3(bt) = 0;
imr = cat(3,im1,im2,im3);
% display
figure; imshow(imr, []);

[Image Analyst]

Try adjusting the threshold in imbinarize().

[Syafri Arlis]

thanks you, to find the automatic threshold value in imbinarize, can you help us

[Image Analyst]

@Syafri Arlis if you use 'adaptive' like @yanqi liu used, then, unfortunately, imbinarize() does not return a matrix of thresholds that were used at every pixel.

If you use the global option, the single threshold is the same as what graythresh() returns.

[Syafri Arlis]

@Image Analyst can you help other than the imbinarize function, what function can we use to retrieve the required object like in the previous image, thank you

[Image Analyst]

@Syafri Arlis, not sure what you're asking. Both @yanqi liu and I attached source code for you in our answers. What's wrong with the imbinarize() function?

Did you actually look at the source code in my answer? If so you'll see I don't use imbinarize like @yanqi liu and I do this:

%===========================================================================================================
% Threshold the image to make a binary image.
thresholdValue = 260;
binaryImage = grayImage > thresholdValue;

Go back up to my posted answer and download the attached files, try them, then adapt them to your code.

[Syafri Arlis]

can you help me, provide an example script to change the color intensity value of a predetermined coordinate point, like the picture

[yanqi liu]

yes，sir，may be use

clc; clear all; close all;
img = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/816674/image.png');
img2 = imcrop(img, [37    63   169   185]);
bw = imclearborder(im2bw(img2));
bw1 = imdilate(bwareafilt(bw,1), strel('disk', 3));
bw2 = imdilate(bw, strel('disk', 3));
% set it to black
img3 = img2;
img3r = img3(:,:,1);img3g = img3(:,:,2);img3b = img3(:,:,3);
img3r(bw1) = 0;img3g(bw1) = 0;img3b(bw1) = 0;
img3 = cat(3,img3r,img3g,img3b);
figure; montage({img2,img3}, 'Size', [1 2], 'BackgroundColor', 'w', 'BorderSize', [2 2]);

% set it all to black
img4 = img2;
img4r = img4(:,:,1);img4g = img4(:,:,2);img4b = img4(:,:,3);
img4r(bw2) = 0;img4g(bw2) = 0;img4b(bw2) = 0;
img4 = cat(3,img4r,img4g,img4b);
figure; montage({img2,img3,img4}, 'Size', [1 3], 'BackgroundColor', 'w', 'BorderSize', [2 2]);

[Image Analyst]

@Syafri Arlis, Not sure what you call white, but let's say it's anything above a certain threshold, like 200.

% Define a threshold that defines what "white" is
threshold = 200;
% Get a mask of where the white pixels are in the entire image.
mask = grayImage >= threshold;
% Define rows and columns of dashed red box.
row1 = 1;
row2 = 150; % Whatever....
col1 = 250; % Whatever...
col2 = size(grayImage, 2);
% Make sure mask is false outside the red box 
% (just in case there might be some white pixels there,
% though there aren't any for this particular image that is shown).
mask(:, 1 : (col1 - 1)) = false; % Erase left half.
mask(row2 : end, :) = false; % Erase bottom half.
% Turn white into black (0) but only inside the red box.
grayImage(mask) = 0;

[Syafri Arlis]

@yanqi liu, @Image Analyst for the image has been converted to binary (0,1) and we are trying to remove the marked area from value 0 to 1, can you help me?, thank you

[yanqi liu]

yes，sir，may be use

clc; clear all; close all;
img = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/816764/image.jpeg');
bw = im2bw(img);
bw1 = imfill(bw, 'holes');
bw2 = bw1;
bw2(round(size(bw2,1)/2):end,:)=bw(round(size(bw2,1)/2):end,:);
figure; montage({bw,bw1,bw2}, 'Size', [1 3], 'BackgroundColor', 'r', 'BorderSize', [2 2]);

[Syafri Arlis]

@yanqi liu Thank you for the help

[Syafri Arlis]

@yanqi liu We are trying to separate the left and right areas, according to the attached image, can you help me?, thank you

[yanqi liu]

clc; clear all; close all;
img = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/853405/image.png');
im = imcrop(img, [24    88   379   219]);
im = rgb2gray(im);
bw = im2bw(im);
bw = imclose(bw, strel('disk', 19));
[L,num] = bwlabel(bw);
stats = regionprops(L);
rects = cat(1, stats.BoundingBox);
[~,ind] = sort(rects, 1);
im1 = im; im2 = im;
im1(L~=ind(1)) = 0; im2(L~=ind(2)) = 0;
figure; 
subplot(1,3,1);  imshow(im, []);
subplot(1,3,2);  imshow(im1, []); title('left');
subplot(1,3,3);  imshow(im2, []); title('right');